orxtra

orxtra is a Python framework for AI agent orchestration that runs every unit of work as a nested task with entry and exit checks, a USD budget ceiling, and a PostgreSQL event trace. It is built for developers and for the agents themselves, who drive it through a CLI, an MCP server, or the Python API. Verification is structural rather than advisory: no tool call can happen outside a task, and a failed exit check sends the task back to the agent instead of sending its result on to the caller.

The problem

AI agent orchestration today is unstructured. Agents spawn agents, there are no verification boundaries, no budget enforcement, and no systematic way to know if work was done correctly. Free-form agent delegation is the goto of AI workflows — it produces results, but they are fragile, opaque, and unverifiable.

The solution

orxtra applies structured programming to AI workflows. Every piece of work is a task with explicit entry conditions (pre-checks), exit conditions (post-checks), and a budget ceiling. Tasks nest recursively. Failure propagates up the hierarchy. The system verifies work at every boundary — not after the fact, but as a structural guarantee.

What makes it different

Modules

Six more foundation modules (protocols, secrets, write-safety, notepad, session, services) are installed as dependencies when you need them.

Quick start

One-shot LLM call, no orchestration:

from orxtra.services import ask

result = await ask(
    prompt="Summarize this document.",
    provider_type="anthropic",
    model="claude-sonnet-4-20250514",
    api_key="sk-...",
)

Structured workflow with verification:

from orxtra.services import start_run, RunConfig

run_id = await start_run(
    pool=pg_pool,
    workflow_path="workflows/review.toml",
    config=RunConfig(
        provider_type="anthropic",
        model="claude-sonnet-4-20250514",
        api_key="sk-...",
        autonomy_level="medium",
    ),
)

Examples

• basic_agent.toml — minimal agent definition
• simple_workflow.toml — workflow with dependencies and post-checks
• categories.toml — multi-provider model routing
• coder_agent.toml — agent with exec capabilities

CLAUDE.md — full technical reference | CHANGELOG.md — release history